Thread by @lxeagle17

🚨GEORGIA UPDATE 12/24 (TOTAL + VBM + IN-PERSON)🚨

Total Votes: 2,062,902 (+178,300)

Percent of registered voters: 26.7%
White: 55.5% (56.5% in general)
Black: 31.5% (27.7% general)
Hispanic: 2% (2.6% general)
Asian: 2.3% (2.6% general)

COUNTY TURNOUT RATES (shown as % of registered voters):

State Average: 26.7%
DeKalb: 31.5%
Cobb: 27%
Fulton: 30%
Gwinnett: 27.5%
Chatham: 18.9%
Muscogee: 23.7%
Columbia: 27.5%
Houston: 28.1%
Baldwin: 16.2%
Fayette: 33.8%
Hall: 24.1%
Clayton: 25.4%
Forsyth: 33.3%

🚨IN-PERSON EARLY VOTES🚨

Total Votes: 1,341,150 (+135,667)
Percent of registered voters: 17.3%
White: 55.9% (58.2% in general)
Black: 31.5% (26.4% general)
Hispanic: 2.1% (2.6% general)
Asian: 1.9% (2.0% general)

COUNTY TURNOUT RATES (shown as % of registered voters):

State Average: 17.3%
DeKalb: 19.8%
Cobb: 11.8%
Fulton: 21.7%
Gwinnett: 16.9%
Chatham: 11%
Muscogee: 17.6%
Columbia: 18.7%
Houston: 16.4%
Baldwin: 11%
Fayette: 24.6%
Hall: 15.5%
Clayton: 16.7%
Forsyth: 23.5%

🚨VBM DATA🚨

Accepted Votes: 679,209 (+42,546)
White: 54.9% (53% general)
Black: 31.6% (30.3% general)
Return Rate: 53.5%

Requested Votes: 1,347,856 (+7,468)
Request Rate: 17.4%
White: 51.6% (51.2% general)
Black: 32.7% (31.4% general)
66+: 46.1% (40.7% general)

COUNTY REQUEST RATES (shown as % of registered voters):

State Average: 17.5%
DeKalb: 24.6%
Cobb: 25.6%
Fulton: 17.8%
Gwinnett: 19.9%
Chatham: 19.6%
Muscogee: 17.3%
Columbia: 14.2%
Houston: 18.7%
Baldwin: 19.6%
Fayette: 19.3%
Hall: 15.2%
Clayton: 16.6%
Forsyth: 16.3%

Per @Nate_Cohn’s primary history data, the electorate share was just-so-slightly pro-R today, which matches up with the estimate of @joe_gantt from yesterday about what we were expecting. There’s a margin of error here, of course, but it’s what we have for now.

It was a slightly more pro-R day in terms of partisanship than Monday, but with an important caveat — the votes cast today dropped off significantly.

That’s pretty important, because the electorate is currently something like 56-44 D, as analysts like @WinWithJMC and @Nate_Cohn have shown, and its borne out by @joe_gantt and @adincgor’s analysis as well.

So it’s now also about the total votes cast also, and the GOP would have liked to see more turnout today to close the gap. The model yesterday (before it was released) had an estimate of Ossoff +1.3, and it’s now Ossoff +1.4, because expected a slightly more pro-GOP day.

Something interesting, per Cohn, is that the total votes cast have been surprisingly getting younger and younger to the point where we’re now level with the general election in terms of the 18-to-29 age demographic’s share and could overtake it! So IDK if the runoff will be older

On a county basis, the Republicans had a pretty good day in educated areas like Forsyth, Fayette, and Cherokee. However, non-college counties like Hall and other areas seem to be lagging a bit behind statewide average still, which is not great...

And that dip is where Democrats come into play — Fulton/DeKalb have shown no indicates of slowing down at all and have been taking up a huge statewide vote share.

Clayton and Hancock, meanwhile, act as bellwethers for the Black turnout, which we see consistently remains high and appears on pace to well exceed November, maybe by 2%.

If that is the case, you can then also expect the margins in Fulton and DeKalb to be better for Ossoff than what he had in November, which *also* goes further into wiping out Perdue’s November advantage.

Anyways, there is a clear path back for the GOP, and it relies heavily on in-person turnout surging next week. Lean is NOT safe, and this race is not over. If they have a surge next week and do well on election day, they can win. But they're behind November rates right now.

So here's what we have currently:

Model rating from @ADincgor based on county turnout:

Ossoff +1.4, Lean Democrat.

docs.google.com/spreadsheets/d…

It's lean because at tossup, I have to be able to construct an argument for either side being favored. I can't do that right now, so it's lean Democrat.

Again, lean is *not* safe. Thank you for coming to my TED talk.

and thanks to @ElectProject, @ADincgor, and @joe__gantt for the data, to Ali for the model, and to Ali and Joe for the help analyzing all of this.

Slight correction: 178,210 ballots is the total net change in votes. For VBM accepted rates, it's +42,543 votes there instead. Lots of different numbers floating around in the spreadsheet; sorry about the small error there, y'all.